Garry Snook thinks risk is overrated. Applying by-the-numbers rigor to every decision he makes, he has built Performance Bicycle Shop into an industry leader

In the warehouse the flow of product is relentless, moving from overhead bins that reach the 22-foot-high ceiling and converging on a central conveyor jammed with packing boxes. The boxes are filled, sealed, and loaded into a seemingly endless procession of tractor trailers backed up to wide-open doors at the back of the building.

The man at the center of all this motion wears a white shirt and red silk tie. Of medium height, with dark, darting eyes, he might be mistaken for an IRS agent on the prowl or a visiting banker (which in a former life he was). "I saw more companies get into trouble," he is now saying here amidst the hum of the warehouse. "They would build space before they had sales. Then they'd be sitting with a huge building, a fancy computer system -- all these fixed costs -- and not enough revenues." He shakes his head. "We avoid that. We always build late. We wait till the place is bursting at the seams."

To buttress his point he indicates a spot in the floor where the texture of the concrete suddenly changes. Small divides like this one can be found all around the building. Worn, old carpet suddenly gives way to new, higher-napped stuff. Smooth concrete yields to rough. Like telling layers of sediment, they signal moments in time when a wall was pushed out and the space expanded to accommodate one more surge in a fast-growing operation.

In 1982 Garry Snook scraped together $25,000 in savings and started a small mail-order operation selling cycling parts and accessories out of his basement in Chapel Hill, N.C. Today Performance Bicycle Shop is a $50-million-a-year business, the largest single retailer of cycling gear in the United States. The Performance empire is a real tree killer, cranking out 10 million catalogs a year. It also includes 14 retail shops that on average do five times the volume of the typical U.S. bike shop.

In looking at the big Performance payoff it is tempting to assume that Garry Snook is a man who bicycles 150 miles a week and favors cleanly crafted cranks and derailleurs in his dreams -- yet another high-rolling entrepreneur who turned a passion into a profit.

Not so.

Garry Snook, age 45, Duke M.B.A. 1981, rarely rides bicycles, preferring instead to get around in a monster Mercedes-Benz. He approaches his business with a cool eye cast not on the wild blue yonder but on the warehouse floor. Snook believes business is more science than art and that it can, for the most part, be quantified. He thus poses hard economic questions of virtually every decision he makes, focusing first on his downside risk, then trying to calculate the upside. If the payoff seems a large enough multiple of the cost, he tests the market. If the test proves out, then he makes his move.

Garry Snook comes to his work with a competitor's sense of urgency. He loves business most for the puzzle it presents, something he labels "a complex, nontrivial game." In the case of Performance the game has most resembled chess played by a master: a series of incremental, very calculated moves woven together as the game plays out in a seamless web encircling the prize.

Move #1: Finding the Business. Garry Snook's methodical approach to building a company reaches back to Performance's very start, Christmas of 1981. He knew he wanted to run his own show. What kind, exactly, was secondary to the constraints he faced. They would define his business.

His major constraint was, predictably, money. Snook had $25,000 -- his life savings -- to play with. He needed a business in which capital requirements were low and the opportunity to leverage other resources existed. Then his brother, an avid cyclist, told him about the poor service and choice of goods offered by the typical bike shop. When Snook took a harder look he noted that cycling products "were fairly expensive and very UPS-able." The fitness boom was in high gear, and cycling's popularity was notably on the rise. Moreover, a survey put the average age of committed cyclists at 32, not under 20 as Snook had thought. "These were people with discretionary income," Snook says. "That made a big difference."

Snook went to the library and checked out a book on how to start a mail-order business. He then studied the half a dozen cycling catalogs that did exist. Many ran small space ads in the nether reaches of cycling magazines. Most were poorly produced. Some charged a fee just to send a catalog. Few had 800 numbers. None offered warranties of any substance. They competed on price alone; customer service scarcely existed.

Snook sensed a niche was there to fill. He would do for cycling what L.L. Bean had done for outdoor wear. He would print a good-looking four-color catalog, have an 800 number that would be answered round the clock, and offer quality goods at a fair price. He would offer a 100%-money-back guarantee, no questions asked.

Snook struck a deal with a printer that had been his client when he was in banking, and in April 1982 he mailed 35,000 16-page four-color catalogs. If Snook printed that many and the business tanked, he would only have been broke. If he printed more than 35,000 catalogs and the business failed, he also would have had to file for bankruptcy protection, and that was where he knew he had to draw the line. "I was not willing to lose my reputation," he says.

That first catalog carried as many as 10 items per page, offering the illusion of a broad selection between the covers of a slim book. "I wanted to spread my risk out as much as possible," says Snook. "I also couldn't tell which items would sell, and which would not." He would order five of each item at a time, pay COD, sell what he had, and quickly reorder. "Our suppliers could fill orders in a timely manner because we were so small." The printer also helped out, spreading Snook's payments over 90 days.

Snook projected a 2% response rate and an average order of $45. That would get him into a second printing of another 35,000 catalogs. Sales spurted, however, and Snook printed 100,000 more catalogs. His wife, Sharon, who had delivered their second child just two weeks before the first mailing, helped Snook fill orders and keep records. The phone rang with such insistence they could never be away from the house together. When the weather got warm, Snook installed a bell that would ring outside when the phone rang inside, allowing him at least to get out and cut the grass, if somewhat fitfully. In its first year Performance grossed $700,000.

Move #2: Securing the Niche. In its second full year Performance soared to $3.2 million in sales. Snook had found quite a niche, but now he worried about securing it. The company, he reasoned, was vulnerable. "My competitors were all price marketers," he explains. "They could have outlasted me." Performance had differentiated itself from the pack with customer service, yet cyclists were a price-sensitive lot. Snook knew he had to provide customer service and compete on price.

His research told him that cyclists express little brand preference for soft goods -- apparel and accessories. They are less concerned with label than with price and performance. But with hard goods -- bike frames and components -- they exhibit strong brand loyalty; Campagnolo, Shimano, and Suntour cranks and derailleurs have evident cachet. "There was no point in trying to compete with the best," reasons Snook. Moreover, hardware was expensive to produce. So he decided to create a line of apparel bearing the Performance label, a line that would offer the quality and value cyclists looked for. He would start by risking as few resources as possible -- to make sure his hunch was right.

In the spring of 1983 Snook called on another former client from his banking days, Pickett Hosiery Mills Inc., in Burlington, N.C. He persuaded Nim Harris, the owner of Pickett, to sell him the tail end of a run of high-quality socks the mill was producing for a major sporting-goods company and to print the Performance logo on the socks. "I couldn't have afforded this if Pickett had to make a special run just for me," Snook recalls.

He paid $1 a pair and ordered just 10 dozen pairs. Some risk. He would sell the socks at $2.25 a pair, comfortably under the $3.50 that identical socks were fetching in retail stores. The socks sold out within a month, and only then did Snook reorder.

At $2.25, what surprised buyers was the quality. Snook had begun to establish the reputation he wanted for the Performance label. He has since been able to expand the Performance line of socks to include seven varieties. One even comes with a three-year, 18,000-mile guarantee. In the past five years Performance has sold 300,000 pairs of socks bearing the company logo. And that was just the beginning.

Since 1983 the company has had a product-development group looking for niches into which Performance can fit its private brand of products, while, conversely, avoiding areas in which, because of existing quality or the high cost of entry, it would be foolish to compete. Only about 20% of the hardware Performance carries bears the company name. The rest is made by name manufacturers. In soft goods, the reverse is true; 80% carry the Performance label. And for good reason.

In a July 1989 survey of the readers of Bicycling magazine, 44.4% responded that they owned Performance apparel -- a percentage far higher than that of any other brand. While Performance sportswear accounts for 80% of the clothing advertised in the company's catalog, it rings up 90% of the apparel sales.

Move #3: Controlling Production. By January 31, 1985, Performance's annual sales had climbed to just over $7 million. Garry Snook had a real company on his hands, and what he sensed about real companies is that they learn how to control their destinies -- or they perish. A key step toward that goal, Snook believed, was to start bringing manufacturing in-house.

Snook learned that the costly and painful way when he began having delivery problems on a staple item, cycling shorts, which were supplied by an Italian manufacturer. He was then struck by the absurdity of having a company based in textile country and ordering product from half a world away. If he could reduce the number of links in the chain, maybe someday he'd see $20 million.

Driving from Chapel Hill to Burlington, 25 miles to the west, you pass from one era of the South to another, from a well-scrubbed college town with its bookstores and coffeehouses, to one-story cinder-block churches and begrimed brick mills -- relics evocative of a time before the global textile wars broke out. Pickett Hosiery Mills in Burlington covers 80,000 square feet, its hardwood floors worn smooth and supple by the tread of generations earning steady paychecks. Here amid a fluorescent haze and the high-pitched whine of sewing machines sit rows of women dwarfed by garments piled high beside them.

In September 1984 Garry Snook walked into that scene, looking not just for more socks but for a piece of the space. Nim Harris, long on space, gladly rented him 1,000 square feet for all of $500 a month, utilities included. The lease ran from month to month. Snook bought three sewing machines at a bankruptcy auction for $200 apiece. He hired three seamstresses. He ordered 400 yards of Lycra from a local job shop at $5 a yard. That would yield about 500 pairs of shorts.

Snook had been selling his Italian-made shorts for $24.95 and $26.95. He now was able to sell all his made-in-the-USA shorts bearing the Performance label at $19.95. He had eliminated a middleman, and an unreliable one at that. The shorts continue to sell at 45,000 pairs a year, having long since reinforced the image of the Performance brand as one of quality. Again, as with the socks, the success of a single product, well-made and well-priced, provided Snook the leverage to expand his line. Performance today sells 11 types of shorts.

Looking back on that first homemade pair, Snook notes, "Those shorts really built the business." His cut-and-sew operation has mushroomed into a full-fledged subsidiary, Nova Designs, which now employs 100 people and last year rang up $8 million in sales to Performance. Nova makes more than half the apparel sold under the Performance label. The balance, more labor-intensive, is done overseas.

When Snook brought manufacturing in-house he gained flexibility. Performance could move faster on an idea, order low minimums, and reorder and reproduce items quickly if a test idea proved hot. Observes Snook: "If we had to rely on a foreign supplier, we might have to wait 12 to 16 weeks for redelivery. By then the season could be over."

MOVE #4: Pushing into Retail. By 1986 Snook felt his company bumping up against the edges of its niche. There were only so many people you could mail cycling catalogs to. And he worried about the vulnerability of the mail-order business. What if postal rates spiked? What if he had to charge sales tax? What if he could no longer accept credit cards over the phone? Snook concluded that if he was going to be a long-term player in the bike business, a thriving mail-order operation would not be enough. It was time to push into retail.

Snook's market research confirmed his impulse. The average bike shop was often run by enthusiasts with scant business experience. The presentation was hit-or-miss, the selection spotty, the ambience -- scuffed linoleum and pegboard walls -- a turnoff. Most customers walked out of bike shops unhappy with the service.

Snook thought once again about the typical 32-year-old bicyclist with discretionary income. "I think this customer wanted something a little more upscale," says Snook, who sensed another chance to differentiate Performance.

Snook decided to open 6,000-square-foot stores -- which would be more than twice the size of the average bike shop (2,500 square feet). He would devote 40% of the space to clothing, while the average bike shop devoted only 5% to those items. That would give the stores not only a brighter, more diverse look but fatter margins, too. Bicycles account for 50% of sales in the average bike shop. At the same time they have the lowest margins (about 30%). Soft goods, which Snook would feature, have the highest margins in the business (about 50%). Snook would bring the hardware out from hiding in the back room to where customers -- who love components -- could touch and feel the finished product. He would detail the stores with cherry trim, carpets, and parquet flooring.

The catalog operation, by this point, had sold to thousands of customers, providing a valuable database. "I know where more than a million cyclists live," Snook says. Catalog sales were strongest in three areas: California, Colorado, and around Washington, D.C. He chose Colorado after a marketing consultant found that if Performance's prices were 10% less, people on average would drive 23 minutes to his store. That was key: since Performance's prices averaged 20% less than other retailers, Snook could consider locations other than Denver. He leaned toward Boulder, about 25 miles to the northwest, a bike-mad town that had nearly seven times as many cyclists per capita as the average U.S. city. Performance had to be there.

Going in, Snook knew his lease would be his single biggest cost, since the typical lease for the type of space Performance sought is 10 years at $100,000 a year -- $1 million. For Snook that was too big a bite. Taking advantage of the Mountain region's rocky economy, he chopped the term in half, then negotiated a "kick out" clause at the end of the second year. If the shop failed to hit certain revenue targets, Snook could bail out.

Snook knew that the average bike shop in major metro markets grossed $450,000 yearly, but he was confident he could do $800,000 his first year in Boulder. That was break-even. "My space was larger, my selection broader, and my presentation better, and my prices were lower," Snook recalls. Boulder already had seven stores, averaging more than $500,000 in sales. Snook felt sure he could take some of that market because his widely distributed catalog would presell the Performance retail outlet, and word of mouth in the tightly knit cycling community would do the rest.

In August 1986 the Boulder store opened, and the word got out fast. Sales hit $1.3 million in year one, and annual growth has averaged 10% since then. A year later Snook opened a store in Denver that also did $1.3 million in sales its first year. In March 1989 he opened a store in Lakewood, a Denver suburb, where the first-year gross again exceeded $1 million. (That store is in a shopping center anchored by a Safeway. Snook negotiated a clause that stated that if Safeway ever left the site, Performance would be free to break its lease.)

Performance now has 14 stores, accounting for 40% of the company's sales. Less than 5% of all bike shops in the United States gross more than $1 million a year. All of Performance's stores do at least that. Its store in Redwood City, Calif., grossed $1.9 million in just its second year.

That kind of success, you might think, could get mainlined to Garry Snook's cerebral cortex and move him to madly open new retail outlets from Jacksonville to Juneau. Performance, in fact, is going slower this year than projected, opening four stores in 1991 instead of six. Inventory shrinkage is too high to suit Snook. The shrinkage rate last year was 2.5%, against an industry average for specialty retailing of 1.7%. "We recently got it down to 1.27%," says Snook proudly. Labor, at 13% of sales, was also a problem. "That's way too high. We've got it down to 9.5%," he says, reciting these numbers as if they were tattooed on his brain.

Move #5: Synergy. In looking at Performance it is safe to say Garry Snook has found the right leverage points. The multiples of his original investments have turned out to be huge. Think what that first 10 dozen pairs of socks cost Snook compared with the yield. Think what has come from that leased 1,000-square-foot space at Pickett Hosiery Mills. Consider that a $25,000 investment has led to a $30-million mail-order operation. You could argue that Snook's risk on the retail side amounted to not much more than $250,000 -- a year's rent in Boulder, improvements on the space, labor. (He could always sell the inventory through the catalog.) The 14 stores now yield $20 million in annual sales. By 1995 Performance expects to have 50 stores.

And now comes the easy part.

Performance's future growth will be bolstered by a host of synergies found throughout the business. The catalog has become an efficient means of advertising the retail stores and identifying where serious cyclists live. The stores, in turn, have yielded new catalog customers. Snook is currently considering a "B" store concept in smaller population centers near his existing stores. These will be 3,000-square-foot spaces stocked with the better-selling inventory from the big stores.

Last May Snook opened a seasonal store in Aspen, open from May to October and measuring only 1,200 square feet. The space is sublet from a ski-rental outfit, eager to be off the high-rent hook for the five off-season months of the year. Snook leases the store for $10,000 for the five months and intends to rent a lot of mountain bikes. In the fall he can close up shop and send his inventory back to Boulder. He thinks he can gross at least $300,000 in Aspen. If the idea takes off, he obviously can roll it out in a number of similar locales: Stowe, Vt.; Jackson Hole, Wyo.; Taos, N.M.; Park City, Utah.

Today, nine years after he entered the game, Garry Snook has built a business with formidable barriers to entry. The Performance label, having appeared on a wide range of goods in Snook's stores and catalogs, is recognized by customers as standing for quality at a reasonable price. Snook now mails out almost 10 million cycling catalogs a year, far more than any other cycling publication. If someone wants to compete with Performance, he or she must be prepared to mail in that kind of volume and to the right people. That would take both a lot of money and a lot of know-how. Says Snook: "We are a big fish in a little pond. I'd have to say the easiest way to compete with us would be to buy us."

Driving into Burlington one day to look things over at his cut-and-sew operation, Snook motions out the window of his Mercedes at a decrepit storefront next to a Chinese restaurant, where he once leased production space. He shakes his head, disbelieving that once Nova could fit into that building. Then his mind abruptly turns back to solving the ongoing puzzle. He starts running through the numbers on his Aspen idea, marveling at how low his break-even was; he talks about recent test mailings he has done of catalogs featuring ski and scuba gear. A golf catalog will be next. "Do you realize that's a $5.7-billion industry?" he says, wheeling into the parking lot at Nova. He has found a golf-club source in Taiwan that will sell him a minimum of 100 sets. "That's nothing. That's so low."

Nine years and one terrific business later, Garry Snook is still running through the numbers, looking for the next opportunity. For him, the game endures -- complex, nontrivial, and very lucrative.